Skip to content
Friday, November 13, 2026 · Today’s edition, 6:30 a.m. ETFri, Nov 13 · 6:30 a.m. editionSign inMy Desk 0EditionsSign in

Arnott puts top-five U.S. concentration at 30%

Rob Arnott says $1 trillion a year of AI capex implies $150 billion to $200 billion of annual write-offs on a five-to-seven-year depreciation schedule.

At a glance

  1. Arnott, founder of Research Affiliates and now chairman of Syzygy Asset Management, made the comparison to Cinthia Murphy, TMX VettaFi's director of research, and said concentration is a consequence of expectations.
  2. For an advisor holding the S&P 500 or Russell 1000 as a core equity sleeve, the top five names already sit at benchmark weight before any manager makes a selection.
  3. Arnott put the annual write-off cost of the AI buildout at $150 billion to $200 billion, based on $1 trillion a year of capex on chips and data centers depreciated over five to seven years.

The top five stocks in broad U.S. indices now account for roughly 30% of the entire U.S. stock market, Rob Arnott said on Bloomberg's ETF IQ, versus 27% for the top five in Israel.

Arnott, founder of Research Affiliates and now chairman of Syzygy Asset Management, made the comparison to Cinthia Murphy, TMX VettaFi's director of research, and said concentration is a consequence of expectations. Those expectations extrapolate the spectacular growth of the largest businesses, he said, and that extrapolation is dangerous. He cautioned that today's concentration rests on an unsustainable extrapolation of past performance, the kind of reasoning that conventional market wisdom supports by invoking precedents such as the 1992 Olympic "Dream Team" to justify megacap dominance. ETF Trends reported on the segment.

For an advisor holding the S&P 500 or Russell 1000 as a core equity sleeve, the top five names already sit at benchmark weight before any manager makes a selection. Picking a cap-weighted tracker or a smart-beta or fundamental index is a decision about the weighting rule. ETF Trends calls smart beta strategies an antidote to the concentrated exposure that market-cap-weighted indices can carry, and said the segment also covered the evolution of fundamental indexing.

Arnott put the annual write-off cost of the AI buildout at $150 billion to $200 billion, based on $1 trillion a year of capex on chips and data centers depreciated over five to seven years. ETF Trends said that accelerated depreciation schedules create severe earnings vulnerabilities at the technology companies funding the buildout.

ETF Trends attributes part of the risk to the mechanical influx of passive capital into standard market-cap-weighted vehicles, which it says drives a wedge between index constituents and non-members regardless of underlying cash flows. The publication described the U.S. equity market as grappling with unprecedented concentration risk.

The arithmetic behind Arnott's range is division: $1 trillion spread over seven years is about $143 billion a year, and over five years $200 billion. The depreciation schedule moves the annual write-off across that band.

Reported by the Passive & Indexing desk from ETF Trends and Bloomberg ETF IQ. Published . Corrections

Related

ETFs

RAFI pitches fundamental indexing as a third option to cap weight and equal weight

RAFI Indices puts the weighted average market cap of its fundamental index at $769 billion, between the $1.39 trillion of a cap-weighted portfolio and the $127 billion of equal weight.

ETFs

RAFI's $200 billion anti-cap-weight business lands at TMX

Rob Arnott's fundamental indexing now sits behind TMX VettaFi's push for $1 trillion in index assets.

ETFs

Research Affiliates ETF holds deleted stocks up to five years

NIXT rebalances deletions from top 500 and top 1,000 market-cap indices to equal weight each year, tracking the Research Affiliates Deletions Index.

ETFs

Direxion's Nvidia bull and bear ETFs face Nov. 17 Nvidia earnings

NVDU seeks 200% of Nvidia's daily return and NVDD seeks the daily inverse, unleveraged.

ETFs

AMLP and ENFR raise Q4 distributions, MLPA cuts its payout

AMLP pays $1.04 a share on Nov. 17, up 4% from a year earlier, while MLPA cuts its payout to $0.83, down 18.6%.

ETF Dailyfrom Private Wealth Daily

Flows, launches, fee moves and the funds advisors are adding to models. Daily at 6:55 a.m. ET.

Free. Every morning. Unsubscribe from any edition in one click.

From other sections

Arnott puts top-five U.S. concentration at 30%